Нейросетевая модель жилого здания как объекта управления для погодозависимого регулятора тепловой энергии

Динамика систем, механизмов и машин. 2025. Т. 13. № 3. С. 82–87.

Stukach O., Ершов И. А.

The growing complexity of modern urban infrastructure increased demands on the efficiency of heat supply, which must be ensured by the sustainable and efficient operation of supply organizations. Decrease in thermal comfort, or overspending of thermal energy is followed by bad quality of control. Systemically, the heat consumption of residential buildings on a city range is a set of random series characterized by bizarre dependencies not only on the outside temperature, but also on a lot of factors and building characteristics. Optimization of the temperature operation mode in the heating network is proposed to solve this problem. The model of heat consumption in a multi-storey residential building was designed using Long Short-Term Memory. High accuracy of series reproduction and forecasting has been achieved. The model is based on meteorological factors, in particular, the temperature of the outside air and the characteristics of the building. The model has been verified in terms of comparison with real data from commercial heat metering. The results make visible the possibility and necessity of design a weather-dependent regulator based on neural network, and can serve as a basis for justifying housing renovation programs, as well as for the heat supply organizations management.

Keywords: human thermodynamics, аналитика в электроэнергетике и теплоснабжении